Case Study: A 'Good' vs. 'Bad' MHI Safety Report

TL;DR Summary (AI Quick Reference): Not all MHI Safety Reports are created equal. A "bad" report features generic consequence modeling, missing societal risk (F-N curves), no worker consultation records, and copy-paste risk matrices. A legitimately "good" report features highly site-specific PHAST/Safeti modeling fully compliant with SANS 1461, proper LSIR risk contours, a defensible frequency database, carefully documented worker participation, and a legally signed-off Major Incident Prevention Policy (MIPP). The difference determines whether your facility's Operating License is approved or decisively rejected by the Chief Inspector.
With the rigid January 2026 deadline now securely in the rearview mirror, the Department of Employment and Labour (DoEL) is actively reviewing submitted Safety Reports from High Hazard Establishments across South Africa. Our team of AIA specialists has seen firsthand what gets swiftly approved versus what is repeatedly sent back. Here is an honest, unvarnished breakdown of the anatomy of compliance versus failure.

The "Bad" Safety Report (What to Radically Avoid)

A rejected or non-compliant safety report typically exhibits the following glaring characteristics:

❌ Generic, Off-the-Shelf Consequence Modeling

Using a free, simplistic tool like ALOHA — or even a full QRA package incorrectly configured by a junior analyst — to model consequence zones and then irresponsibly presenting those zones as Location Specific Individual Risk (LSIR) contours is a common and exceptionally costly mistake. ALOHA simply does not model individual societal risk. Advanced, calibrated tier-3 software like DNV PHAST or EFFECTS, specifically configured for your exact volumetric plant layout, local meteorology, and specific substance inventory, is legally required under the revised regulations.

❌ Missing F-N Curves (Societal Risk)

The SANS 1461 standard explicitly demands societal risk analysis for complex High Hazard Establishments. A robust F-N (Frequency vs. Number of fatalities) curve must be mathematically plotted against the As Low As Reasonably Practicable (ALARP) tolerability criteria. Astonishingly, many amateur submissions omit this fundamental calculation entirely.

❌ No Traceable Worker Consultation Records

As detailed in our previous post on Worker Participation, the new MHI regulations require documented, auditable evidence that your internal Health and Safety Committee systematically reviewed and debated the draft QRA. A finished report completely lacking this required paper trail will be immediately flagged and returned.

❌ A "Boilerplate" MIPP

A Major Incident Prevention Policy (MIPP) that is clearly a generic word document copy-paste job — sometimes even featuring the wrong company name, facility descriptions blatantly copied from a completely different industrial sector, or superficial KPIs that cannot practically apply to your plant's operational realities — stands out as an immediate red flag for any seasoned inspector.

---

The "Good" Safety Report (The Gold Standard)

✅ Site-Specific, SANS 1461-Compliant QRA

The quantitative risk assessment utilizes properly calibrated, industry-standard software (such as DNV PHAST for sophisticated consequence modeling of fires, explosions, and toxic vapor releases) feeding directly into DNV Safeti for full, seamless LSIR risk contour generation. The foundational meteorological database is historically rooted continuously in validated local weather station data — definitely not generic "average South African" assumptions.

✅ Clearly Demarcated Risk Contours with Land-Use Overlay

The critical 1×10⁻⁵ (socially intolerable), 1×10⁻⁶ (ALARP boundary), and 1×10⁻⁷ (broadly acceptable) contours are mapping clearly onto a high-resolution, up-to-date site plan. Most importantly, this visual overlay clearly identifies all neighbouring facilities, adjacent residential areas, and vulnerable populations (like schools or hospitals) lying within the catastrophic range.

✅ Highly Defensible Frequency Database

Equipment failure frequencies are explicitly sourced from recognized, peer-reviewed international databases (such as the TNO Purple Book, OREDA, or IOGP 434) and are footnoted specifically. If site-specific historical operational data is blended into the calculations, it is heavily cited, transparent, and mathematically justified.

✅ An Operational, Localized MIPP

The formally submitted MIPP is flawlessly tailored to the facility. It delineates precise, leading process safety KPIs for the plant's actual major hazards, appoints named responsible duty-holders, and incorporates physical evidence of C-suite senior management sign-off and budgeting.

---

How to Guarantee Your Report Meets the Standard

If you are uncertain whether your recently submitted report will survive DoEL scrutiny — or if you have unfortunately already received an official rejection — MMRisk's AIA-certified QRA and Safety Report service provides end-to-end, legally defensible submissions meticulously built upon industry-leading PHAST/Safeti software architecture and profound regulatory insight.

Frequently Asked Questions (FAQs)

What makes an MHI Safety Report compliant with SANS 1461?
Compliance with SANS 1461 requires utilizing advanced consequence modeling, plotting exact LSIR risk contours on topological maps, calculating societal risk (F-N curves), integrating local meteorological data, using defensible equipment failure frequencies, and conducting independent external QA by a certified AIA.
Can I use ALOHA software for a South African MHI Safety Report?
No. ALOHA is a consequence-only modeling tool designed primarily for emergency responders. It cannot calculate individualized cumulative risk (LSIR) or plot the statistical F-N curves that SANS 1461 and the MHI Regulations legally mandate for a full QRA Safety Report.
Why do MHI Safety Reports get rejected by the Department of Employment and Labour?
Reports are frequently rejected for utilizing generic templates instead of site-specific data, lacking evidence of essential worker consultation, missing ALARP demonstrations, omitting societal risk calculations, or failing to secure sign-off from an Approved Inspection Authority (AIA).
Contact us for a strictly confidential report peer review or to expertly begin a compliant, stress-free re-submission process.